Google Now Uses BERT Models to Organize Top Stories in Search

Google has announced its new organization of top stories in Search. The search engine now uses BERT models along with machine learning to create top stories carousels on the SERP.

According to Google, people depend on Search to find all types of information. And it has continuously improved to provide quick access to information and help users find the most relevant topics.

Now the search engine company is using a similar approach for top stories on Google Search.

In its blog announcement, product manager at Google Search, Duncan Osborn wrote:

“Now, when there are multiple stories related to your search, we’ll also organize the results by the story. So it’s easier to understand what’s most relevant, and you can make a more informed decision on which specific articles to explore.”

Along with recent coverage, the new structure considers the content’s quality, as well as more diverse sources. The goal is to select the content that provides more context and perspective to daily news.

Using BERT Models and Machine Learning Technique to Organize Top Stories

Google said it had to develop new story-understanding technology to power the experience. The tech maps the people, places, and things involved in a news story then finds the connection.

Google also uses machine learning techniques and BERT models to examine related articles and cluster them into groups. That way, people navigating search results for new stories will find the most relevant content.

The announcement reads:

For example, if you search for “NASA news,” you might see results grouped under the distinct news stories “NASA adds five companies to moon bid” and “NASA detects water vapor on Jupiter’s moon Europa,” along with additional results under “Also in the news.”

Finally, top stories carousels will now feature essential information such as related opinion pieces and notable quotes. By providing access to a more rounded view of a news story, the search engine wants to give searchers various angles to explore.

“We hope these updates help people engage with the most relevant information and build a deeper understanding of the world around them,” Osborn said.

The updates are available on mobile devices in the U.S. English search result. And the search engine company intends to roll out more languages to other locations in the coming months.